May 30, 2007
VOTING: TAKEN IN BY HIGH-TECH
(Modified from version published in the Boulder Daily Camera, Sunday April 22, 2007)
Albert A. Bartlett
The New York Times observed that the electronic voting systems "could end up undermining democracy by producing unreliable election results that cannot be truly audited or corrected." (Nov.26, 2006) In order to allow a reliable audit and recount, it is necessary that there be real paper ballots. Computer memories won't do. Data stored in computer memories can be modified by hackers who can probably do their hacking without leaving a trace.
Investigations have been launched to try to discover what went wrong with the high-tech voting systems. The investigators must ask "Why would anyone in his or her right mind have ever thought that the computer-based voting system could function well in the first place?" "Who forgot about Murphy's Law?"
Let's look at the process at the voting places. Using computers at the polling places to verify the status of each voter by interrogating a central database sounds good in theory, but computer glitches can immobilize critical computers and this will leave long lines of voters standing and waiting, unable to vote because their registrations can't be verified. It is better to have the old traditional printed lists of registered voters at each voting place where the verification process is operated by our friends and neighbors. Remember Murphy was an optimist.
Now let's look at the voting itself. The most important requirement is that each voter must register his or her vote by marking a numbered paper ballot that can be counted and recounted quickly and reliably.
In the early days voters marked paper ballots which were counted and tabulated by hand.
Around 70 years ago the IBM people developed punch cards for handling and tabulating data quickly, reliably and repeatedly. Inexpensive and technologically simple hand punches were developed that could be placed in voting booths to punch IBM cards and until recently these were used for voting in Boulder County. The punches could be set up in minutes, they required no electric power or electronic connection to computers. They were easy for the voter to use. They punched IBM cards which the voter could examine for correctness after punching. The punched cards were deposited in a ballot box w hich was easily transported to a central location to be counted and recounted rapidly by machines. This system of punches and IBM cards met every requirement for a good voting system.
An unsuccessful adaptation of this technology was the introduction of IBM cards on which votes were recorded by using a pointed stylus (or a pencil point) to "punch" out perforated holes in the cards. Frequently this left the infamous "hanging chad" which would gum up the machines that counted the ballots. Instead of replacing the stylus with simple mechanical punches that were widely available that would yield clean punched holes with no chad, election officials threw out the whole system, replacing it with computerized systems that meet none of the requirements for a good voting system.
Machines can count IBM cards rapidly, reliably and repeatedly because the cards are small and stiff. The small size and the stiffness allow the machine to place each ballot precisely in a position where it can be read accurately and recorded in a fraction of a second. I was appalled when I went in to vote for the first time with the new Boulder County system. Here we were given large flimsy ballots the size of newspaper sheets which are almost impossible to count at high speed by machine because of both their size and lack of stiffness. This elementary mechanical truth of the near impossib ility of machine reading of large sheets of flimsy paper was understood some 70 years ago when the IBM card system was developed. You cannot have high-speed machine reading if the paper or cards have been folded; hence the admonition that old timers remember, "Do not fold, spindle or mutilate" your IBM cards.
It is sad to see TV images of people in the election offices struggling with large piles of newspaper-size flimsy folded paper ballots, trying to get the ballots out of the envelopes, unfolded, oriented and positioned properly in the electronic reading machine at a rate of perhaps a few a minute. Seventy years ago we had the technology which allowed reliable counting of the IBM cards at the rate of several a second.
Colorado faced the same problem of updating their voting systems as did many other states. A low-cost, low-tech way of overcoming the problem would be to continue the use of IBM cards that the voter either punches with a mechanical punch or marks with a black pen. Either punched or marked cards can be counted quickly, reliably and repeatedly. Colorado, a self-proclaimed center of high-tech, chose the high-cost, high-tech systems and produced a dismal disaster that is a worthy successor to the celebrated, costly and failed high-tech baggage handling system that was initially installed at great expense at the new Denver International Airport.